1. Field of the Invention
The present invention relates to an optical apparatus equipped with a sight line detector, and more particularly to an optical apparatus equipped with a sight line detector for detecting an axis in the direction of a visual point, so-called sight line (sighting axis) at which an observer's (photographer's) eyes are turned through a finder system on an observing plane (focusing plane) on which an object image is formed by an imaging system, making use of a reflected image of eyes obtained when the observer's eyes are illuminated.
2. Related Background Art
Heretofore, a variety of apparatuses for detecting so-called sight line (sighting axis) (e.g. eye camera) have been proposed for determining the position on an observing plane at which an observer's eyes are turned.
For example, in Japanese Patent Application Laid-Open No. 1-274736, a parallel light flux is projected from a light source onto a front eye portion of an observer's eye, and then a sighting axis is obtained from a cornea-reflected image formed by the reflected light from the cornea and an imaging position of the pupil.
Further, the present applicant has proposed an optical apparatus equipped with a sight line detector capable of calibrating a sight line for offsetting the difference in the sight line between observers in Japanese Patent Application Laid-Open No. 3-11492.
FIG. 34 is an explanatory view for the principles of a well known sight line detecting method. In FIG. 34, the numerals 13a, 13b designate light sources such as LED emitting infrared light that is not sensed by the observers. The light sources 13a, 13b are substantially symmetrically disposed in the x direction with respect to the optical axis of a light-receiving lens 12 to illuminate eyes 15 of the observer. A part of the light reflected from the eye 15 is then collected to an image sensor 14 through the light-receiving lens 12.
FIG. 33A is a schematic view showing an eye image projected onto the image sensor 14, while FIG. 33B is a graphic diagram showing an intensity of output signal from the image sensor in FIG. 34. A method of detecting the sight line will now be described with reference to these Figures.
The infrared light emitted from the light source 13b illumintes the cornea 16 of the observer's eye 15. At this time, a cornea-reflected image d (virtual image) is collected through the light-receiving lens 12 and then imaged at a position d' on the image sensor 14.
In the same manner, the infrared light emitted from the light source 13a illuminates the cornea 16 of the eye 15. At this time, a cornea-reflected image e formed by a part of the infrared light reflected on the surface of the cornea 16 is collected through the light-receiving lens 12 and then imaged at a position e' on the image sensor 14. Further, the light flux from the end portions a and b of iris 17 act to make an image of the end portions a and b at positions a' and b' not shown on the image sensor 14 through the light-receiving lens 12.
When a rotational angle of an optical axis of the eye 15 with respect to the optical axis of the light-receiving lens 12 is small, coordinates xc of a central position c of the pupil 19 can be expressed as follows: EQU xc.congruent.(xa+xb)/2
where xa, xb: x coordinates of the end portions a and b respectively of the iris 17.
Further, the x coordinates of the central point of the cornea-reflected image d and e are substantially accorded with the x coordinates xo of a center of curvature O of the cornea 16. Therefore, with the x coordinates of the cornea-reflected image generating positions d and e as xd and xe, a standard distance from a center of curvature O of the cornea 16 to the center C of the pupil 19 as 0C, and a factor for correcting the personal errors for the distance 0C (sight line correcting factor) as A, the rotational angle .theta. of the optical axis 15a of the eye 15 will substantially meet the following relational equation: EQU (A*0C)* SIN.theta..congruent.xc-(xd+xe)/2 (1)
where the mark * stands for multiplication.
Therefore, as shown in FIG. 34, the rotational angle of the optical axis 15a of the eye 15 can be obtained by detecting the position of the characteristic points (cornea reflected images d, e and iris end portions a, b) of the eye 15 projected onto the image sensor 14. The equation (1) can be substituted by the following equation (2): EQU .beta.*(A*0C)* SIN.theta..congruent.(xa'+xb')/2-(xd'+xe')/2(2)
where .beta. stands for magnification determined by the position of the eye 15 with respect to the light-receiving lens 12 and is practically calculated as a function of a space .vertline.xd'-xe'.vertline. of the cornea reflected image.
The rotational angle of the optical axis of the eye 15 can be represented as: EQU .theta..congruent.ARC* SIN {(xc'-xf')/.beta./(A*0C)} (3) EQU where EQU xc'.congruent.(xa'+xb')/2 EQU xf'.congruent.(xd'+xe')/2
The optical axis 15a and the sighting axis of the observer's eye 15 are not accorded. Therefore, if the rotational angle .theta. in the horizontal direction of the optical axis of the observer's eye is calculated, the sight line .theta.x in the horizontal direction of the observer can be obtained by correcting the angle difference .delta. between the optical axis and sighting axis of the observer's eye.
The sight line .theta.x in the horizontal direction of the observer can be expressed as follows: EQU .theta.x=.theta..+-.(B*.delta.) (4)
where B: a factor (sight line correcting factor) for compensating personal error for a correcting angle for the optical axis and the sighting axis of the eye. As to the code ".+-.", if the rotational angle toward the right with respect to the observer is supposed to be positive, it would be "+" when the observer's eye viewing the observing device (finder system) is his left eye, while it would be "-" in the case of the right eye.
Although FIG. 34 illustrates a case of the observer's eye rotating in z-x plane (e.g. horizontal plane), the same detection can be achieved also in the case of it rotating in y-z plane (e.g. vertical plane). In this case, however, the vertical component of the observer's sight line is accorded with the vertical component (not shown) of the optical axis of the eye, such that the sight line .theta.y in the horizontal direction would be expressed as follows: EQU .theta.y=.theta.'
In the case of the observer viewing a finder of an optical apparatus, a position (xn, yn) on a focusing plate can be obtained from the sight line data .theta.x, .theta.y as follows: ##EQU1## where m: a constant depending on the finder optical system of the camera.
In this case, since there are two factors A and B for compensating the personal error in the sight line, it is possible to calculate these factors A and B from the rotational angle of the observer's eye when the observer looks at two visual marks at different positions.
The factors A and B typically correspond to the rotational angle in the horizontal direction of the observer's eye, so the two visual marks provided in the finder of the camera are set to be horizontal with respect to the observer.
Upon determination of the factors A and B and the position on the focusing plate of the sight line of the observer viewing the finder system, the sight line information can be used for the focusing adjustment or the exposure control of the lens.
It is generally not so easy to accurately detect the sight line of the photographer who is viewing the field of view of the finder and to perform photographic control with simplified entire apparatus structure, because of the reasons described below.
For example, if the photographer wears spectacles with high surface reflecting rate, the sight line detecting accuracy would be degraded by the reflected light from the eye when the eye is illuminated by the light flux from a high-lightness object (light source or sunlight) existing in the field of view of the finder, due to the undesirable effect of the ghost generated from the reflected light from the spectacles.
The sight line could not be detected in such cases as: any extremely high-lightness object appears in the field of view of the finder such that the photographer sometimes instantaneously closes his eyes; and the photographer's sight line quickly moves to follow to any high-speed moving object.
In recent years, there have been proposed a variety of cameras incorporating a multiple detecting area-type automatic focus detecting means being capable of automatically detecting focus at a plurality of detecting areas in the field of view of the finder. In this type of camera, the detecting area is previously limited in the photographic screen. Unless the area based on the sight line information from the sight line detecting means is accorded with the detecting area, therefore, the automatic focus detection cannot be carried out on the basis of the sight line information.
In compensating the personal error in the sight line detection of the eyes of the photographers, if any of the previously set correcting values cannot be fit to the photographer, this would cause a detection error leading to an undesirable photographic operation.
There is another method in which a plurality of data for correcting the personal error are stored in the memory means such that many and unspecified persons can operate. This method, however, has such disadvantages that the photographer and the correcting data cannot easily corresponded one to one, and that the memory capacity is restricted.
Further, there is also an alternative method in which data for correcting the personal error in the sight line direction are collected (hereinafter referred to as "calibration"). In this method, it is necessary to newly provide operating means for performing the calibration and the sight line detection, thereby complicating the entire apparatus.
Thus, in cameras having sight line detecting means, the sight line information is sometimes inaccurate or unobtainable. In such a case, the photographer cannot photograph as desired so as not to obtain a desired image.
Meanwhile, the magnitude of the pupil of the observer changes depending on the environmental lightness or his psycological condition. Also, the central position (c) of the pupil and the positions of the Purkinje image (d, e) vary depending on the magnitude of the pupil.
FIG. 35 is an explanatory view for showing a relationship between the pupil diameter rp of the observer i.e., the environmental lightness function, and the position xn at which the eye is turned. The position xn is calculated by the conventional sight line calculating method. When the observer's pupil diameter rp varies in such a manner, the calculated point xp would be changed in accordance therewith even if the observer is viewing the same fixed target. Therefore, the sight line detecting accuracy would be degraded.
For enhancing the detection accuracy, the sight line must be corrected at every change of the observing condition of the observer i.e. the environmental lightness, this would easily disturb the operational property of the camera.